Make the explanation earn its confidence

The Feynman Technique for Studying: A Worked Example

Pick one concept, explain it in your own words, mark the step you cannot justify, and return to an accurate source before revising. The aim is a checked explanation you can use, not a performance that sounds simple. Below is a practical version you can try alone, on paper, without asking someone to sit through a lecture.

By Rain Night Café · Updated 29 October 2026. The example and checklist are our editorial tools. They are not a validation study of the named technique or a guarantee of academic results.

Bring one concept to the free café →

What the technique asks you to do

Birmingham City University’s guide to the Feynman technique describes choosing a topic, explaining it, revisiting gaps and testing the revised explanation. We use that process, without attaching a promised retention percentage to it. Naming a technique after a famous scientist does not make every online claim about it reliable.

The UNC Learning Center’s study guidance also recommends explaining material in your own words, using examples and checking understanding. You can begin with notes, then try again with less support. Keep the source close for checking; pretending not to need help is not part of the exercise.

Choose a concept small enough to inspect

“Explain statistics” invites a vague monologue. “Explain why the mean and median can differ in a small list of numbers” gives you something you can calculate and check. Write the intended listener and the question at the top of a page. An imagined beginner needs enough context to follow your explanation, not baby talk.

Divide the page into four boxes: first explanation, question or gap, source check, revised explanation. The first version does not need to be elegant. Write what you currently think the idea means. Mark borrowed technical terms you cannot yet define, missing reasons and claims that use words such as “always” without stating conditions.

A worked example: mean and median

Suppose your first explanation says: “The mean and median are two ways to get the middle of a list, so they are basically the same.” It sounds approachable, but the phrase “the middle” hides a distinction. Ask a specific follow-up: “If one number becomes much larger, do both measures change in the same way?”

Use the list 2, 3, 4, 5, 6. The sum is 20, so the mean is 4; the median, the middle value in the ordered list, is also 4. This example alone lets the weak explanation survive. Now replace 6 with 26. The sum becomes 40 and the mean becomes 8, while the middle value is still 4.

A revised explanation could say: “For a finite numerical list, the mean is the sum divided by how many values there are. The median is found from the ordered positions. A large change at one end can move the mean while leaving the median unchanged, as in this example.” You have kept the important distinction instead of merely replacing a long word with a shorter one.

Add the boundary your first example omitted: with an even number of values, the usual numerical median uses the mean of the two central values. Check the definitions and conventions required by your course. This is an original arithmetic illustration, not a claim that one summary statistic is always the best choice.

Use a counterexample before polishing the language

A counterexample asks whether your explanation survives a different case. If you say something “always increases,” look for a permitted situation in which it does not. If your explanation depends on a simplified model, name the assumption. The aim is to expose the missing condition before a neat analogy makes it feel settled.

For a history topic, replace the numerical test with an evidence question: what source supports this claim, and what would complicate it? For a coding concept, try a small input that exercises the edge case. For a mathematical proof, preserve the formal argument; a story about the idea cannot replace the steps needed to establish it.

You do not need to invent a difficult exception for every sentence. Concentrate on the claim doing the most work. If you cannot find a trustworthy way to check it, label the uncertainty and take a specific question to your tutor. Leaving one honest question is better than turning a plausible guess into a polished study note.

A checklist for the revised explanation

  • Meaning: Have I defined the key term instead of repeating it?
  • Reason: Can I explain why the important step follows?
  • Example: Does one concrete case show what I mean?
  • Boundary: Have I named an assumption, limit or counterexample?
  • Accuracy: Does it agree with the relevant course source or marking criteria?
  • Next use: Can I answer a new question or apply the idea after this explanation?

Do not score this checklist as a diagnosis of intelligence. It is a way to choose the next revision. A useful next action might be to check one definition, practise one calculation, or ask why a particular assumption is allowed. Avoid rewriting the entire page just because one sentence needs attention.

Keep the explanation, then test it in another form

When the explanation is checked, keep a short version and the source reference. On another day, try a question with the notes covered. Our active recall guide focuses on that question-and-feedback process; the spaced review guide helps decide when to return. You can combine the approaches without treating them as identical.

If an exam expects an essay, formal definition or worked solution, practise that format too. Being able to tell a friendly story about a concept does not automatically prepare you for every assessment. The explanation is a bridge toward the work you need to do, not a substitute for doing it.

Try a quiet, solo version in the café

Writing works when speaking would disturb others. Bring one page, one source and one question. You can use the café’s shared 45/10 rhythm as a boundary, but your exercise can be shorter. Keep sound low or off while working with language. The café does not record an explanation, grade it or match you with a tutor.

Free entry needs no account or camera. Attendance varies; labelled host bots are not human experts. A no-camera shared space can provide optional company while you do your own work, but strangers are not obliged to review it. Café Pass adds optional rooms and personalization, not academic verification. Check the current offer only if that atmosphere would be useful to you.

Try one task in the free café →See what the Café Pass adds

Frequently asked questions

What is the Feynman technique for studying?
It is an explain-and-check approach: choose a small concept, explain it in your own words, identify gaps, consult accurate material and revise the explanation. The useful output is a clearer, checked explanation. Sounding confident or using only simple words does not establish correctness.
Can I use the Feynman technique alone?
Yes. Write for an imagined beginner or explain quietly to yourself, then compare the result with your course material. A partner is optional. You can use the written version in a shared quiet space without speaking or sharing personal information.
Is simplifying everything enough for an exam?
No. Keep required technical terms, assumptions, evidence and formal steps. Use a simple explanation to locate gaps, then practise the type of answer or problem your course requires. A friendly analogy may help introduce an idea, but it can also leave important details out.
Can the café check my explanation or find a tutor?
No. Rain Night Café is an optional ambient place to work, with a shared 45/10 rhythm. It does not grade explanations, verify facts or match tutors. Other visitors are not automatically subject experts, and labelled host bots should not be treated as academic authorities.